cross-sell on the web - pega · 2021. 1. 4. · pegasystems inc. may make improvements and/or...
TRANSCRIPT
STUDENT GUIDE
Cross-Sell on the Web
© Copyright 2020
Pegasystems Inc., Cambridge, MA
All rights reserved.
This document describes products and services of Pegasystems Inc. It may contain trade secrets and proprietary information.
The document and product are protected by copyright and distributed under licenses restricting their use, copying,
distribution, or transmittal in any form without prior written authorization of Pegasystems Inc.
This document is current as of the date of publication only. Changes in the document may be made from time to time at the
discretion of Pegasystems. This document remains the property of Pegasystems and must be returned to it upon request.
This document does not imply any commitment to offer or deliver the products or services provided.
This document may include references to Pegasystems product features that have not been licensed by your company. If you
have questions about whether a particular capability is included in your installation, please consult your Pegasystems service
consultant.
PegaRULES, Process Commander, SmartBPM® and the Pegasystems logo are trademarks or registered trademarks of
Pegasystems Inc. All other product names, logos and symbols may be registered trademarks of their respective owners.
Although Pegasystems Inc. strives for accuracy in its publications, any publication may contain inaccuracies or typographical
errors. This document or Help System could contain technical inaccuracies or typographical errors. Changes are periodically
added to the information herein. Pegasystems Inc. may make improvements and/or changes in the information described
herein at any time.
This document is the property of:
Pegasystems Inc.
1 Rogers Street
Cambridge, MA 02142
Phone: (617) 374-9600
Fax: (617) 374-9620
www.pega.com
Document Name: L-23971-StudentGuide
Date: 04 January 2021
Contents
Next-Best-Action on digital ........................................................................................................... 4
Business use case: Cross-sell on the web ................................................................................ 5
One-to-one customer engagement ........................................................................................... 11
Next-Best-Action paradigm ..................................................................................................... 12
One-to-one customer engagement paradigm ....................................................................... 17
Next-Best-Action Designer ...................................................................................................... 23
Defining and managing customer actions ................................................................................ 30
Action hierarchy ....................................................................................................................... 31
Managing business structure .................................................................................................. 41
Creating actions ........................................................................................................................ 43
Renaming actions ..................................................................................................................... 47
Presenting a single offer on the web ......................................................................................... 50
Real-time containers ................................................................................................................ 51
Creating a real-time container ................................................................................................ 55
Presenting a single offer on the web ...................................................................................... 58
Defining customer engagement policies ................................................................................... 63
Customer engagement policies .............................................................................................. 64
Defining eligibility, applicability, and suitability rules ........................................................... 69
Testing the next-best-action configuration with persona testing ........................................ 74
Avoiding overexposure of actions .............................................................................................. 81
Contact policies ........................................................................................................................ 82
Defining action suppression rules .......................................................................................... 83
Arbitrating between actions ....................................................................................................... 87
Action arbitration ..................................................................................................................... 88
Action prioritization with AI ..................................................................................................... 94
Prioritizing actions with business levers .............................................................................. 102
4
Next-Best-Action on digital
5
Business use case: Cross-sell on the web
Introduction
Pega Customer Decision Hub’s Next-Best-Action Designer lets you configure how you want
the always-on brain to select the best offer for a customer. The best offer is the result of a
series of decisions that are executed in a hierarchical fashion by the brain. Cross-selling on
the web channel will help you improve 1-to-1 customer engagement, drive sales, and
deliver Next-Best-Actions in real-time.
Transcript
This video describes a typical cross-selling use case on the web channel.
U+ is a retail bank. The bank would like to leverage its website as a marketing channel to
improve 1-to-1 customer engagement, drive sales, and deliver Next-Best-Actions in real-
time.
The bank has decided to use the Pega Customer Decision Hub™ to recommend more
relevant banner ads to its customers when they visit their personal portal.
Banner ads are shown on various pages throughout the website.
For example, on the home page, U+ can display a Hero banner at the top of the page,
which is typically a larger image with bigger typeface.
Below that, there is space to display several Tile banners, which are typically smaller.
6
When customers log in to their personal portal, they also see a Tile banner on the ‘Account
overview’ page.
7
8
The main intent of U+ at this stage is to increase their web engagement. This can be
measured by click-through rate. A Click-through is recorded when the customer clicks on
the ‘Learn more’ link.
9
The bank would like to use these banners on the ‘Account overview’ page, to display offers
that are more relevant and likely to receive a positive response.
The offers will be selected by a combination of artificial intelligence (AI) and other business
rules. The AI and business rules are defined in the Pega Customer Decision Hub.
The Pega Customer Decision Hub is the always-on customer brain that acts as a single,
centralized decision authority. The always-on customer brain selects the right offer to be
displayed to each customer who visits the bank’s website.
Next-Best-Action Designer lets you configure how you want the always-on brain to select
the best offer for a customer. The best offer is the result of a series of decisions that are
executed in a hierarchical fashion by the brain.
The bank plans to implement the requirement in multiple phases.
The first phase is a proof-of-concept phase. In this phase, the goal is to display a credit card
offer on the U+ website. This requires getting the basic environment up and running,
setting up the business structure, defining an Action and a Treatment, and enabling
channels and triggers for Next-Best-Action.
As a result of this phase, a credit card offer will be displayed on the ‘Account overview’ page
to all customers who visit the U+ web site. For example, if customer Troy logs in to his
account, the ‘Cash back’ offer is displayed. If another user logs in, they will see the same
offer. However, in practice, more offers should be displayed. Also, not all offers may be
available to a customer for various reasons.
The next phase is to add customer engagement policies. Engagement policies are the set of
conditions such as eligibility, applicability, and suitability that qualify an offer, or a group of
offers for a customer. As a result of engagement rules, customers will see only those offers
that the organization believes they should be exposed to. For example, Troy logs in, he
10
sees the Rewards Card, but for Barbara this is not applicable, so it will never appear;
instead, she sees the Rewards Plus Card.
Too many contact attempts over a short period of time can have a negative impact on a
customer's attitude toward further offers by your company. Therefore, in the next phase,
U+ implements some contact policies using suppression rules, which allow an offer to be
put on hold after a specific number of outcomes. For example, if Troy ignores an ad a few
times, then the ad will no longer be shown to him over a period of time. Instead, his
‘Account overview’ page will show a different ad.
Basically, from a set of all available offers, the choice is narrowed down by engagement
policies. Then the selection is further narrowed down by suppression rules.
After the engagement policies and suppression rules have “whittled down” the total
possible offers to a few, Arbitration is used to choose the top offer based on what is
relevant for the customer right NOW.
Arbitration is the last phase of cross-sell in the U+ web use case.
Arbitration aims at balancing customer relevance with business priorities. Specifically,
Propensity, Context Weighting, Action Value, and Business Levers are given numerical
values. A simple formula is then used to arrive at a prioritization value, which is used to
select the top offer. For example, Troy qualifies for three credit card offers. When he logs
in, he sees the top offer for him, the Standard Card. This offer is the Top 1 because the
priority value is the highest among all other offers.
This video has concluded. It showed you a typical cross-selling use case on the web channel
and its four phases of implementation.
11
One-to-one customer engagement
12
Next-Best-Action paradigm
Introduction
The value of big data and analytics is fully realized when every customer conversation
delivers exactly the right message, the right offer, or the right level of service to provide a
great experience while maximizing the customer’s value to the organization. With Pega
Next-Best-Action, business experts develop decision strategies that combine predictive and
adaptive analytics with traditional business rules to maximize this value.
Transcript
This is your customer. You want him to buy your products, use your services and have a
great experience. And your competitors want the same thing. To compete, you have to
take the right action at every customer touch, ensuring that each conversation delivers
exactly the right message, offer and level of service. You want to provide a great
experience, while maximizing the customer’s value to your organization.
Artificial Intelligence, or AI, can help—if you can get past the hype. Pega has been using AI
to create real business value for years, driving real-time decisions that deliver awesome
engagement on any channel…and improving experiences for over 1.5 billion customers
across the globe.
13
Pega’s omni-channel AI delivers the right action at every customer touch by crunching
millions of data points in real-time. Make an offer, initiate a retention plan, predict a
problem before it happens. Every decision generates the next-best-action for your
customer, and your business.
14
Pega’s AI is built for business people, not scientists or developers. They design
visual decision strategies that combine predictive analytics, algorithms developed through
mining large sets of data, adaptive analytics, machine-learning algorithms that improve
with each interaction, and traditional business rules that allow users to prioritize and
arbitrate between decisions.
Pega uses the strategy to look across all the potential actions you may take with a
customer, make an offer, initiate a retention plan, open a service case, place an ad, and
ensure exactly the right action is taken at every interaction and it works across all channels
to provide a consistent experience in a store, on the phone, on the web, mobile, with the
chat bot, or just some crazy tech that hasn’t even been invented yet.
15
And Pega connects to streams like mobile locations or network events to detect patterns
and drive the Next Best Action proactively. And strategies are completely contextual. Any
change in the customer’s context — a click, a reply, a location change, a Tweet — will trigger
the Next Best Action. So, you can really listen to your customers and act accordingly.
16
Pega’s real-time, omni-channel AI puts the power in your hands, so you can optimize every
customer interaction for experience, and value.
17
One-to-one customer engagement paradigm
Introduction
The optimal outcome of every customer interaction is to provide a great experience while
maximizing the customer’s value to the company. To achieve this, you have to be able to
perform the right action in the right channel at the right moment for each customer. We
call this capability, “1-to-1 Customer Engagement”.
Transcript
In this video, learn about the 1-to-1 Customer Engagement paradigm and how the
principles of Next-Best-Action are implemented using the Pega Customer Decision Hub™.
Customers are more empowered than ever before. As a result, they have very high
expectations of the experiences they receive from their service providers. Their
experiences must make sense within the context of their lives. This means they must be
meaningful, consistent, and personalized across every channel they interact with.
In business, the optimal outcome of every customer interaction is to provide a great
experience while maximizing the customer’s value to the company. To achieve this, you
have to be able to perform the right action in the right channel at the right moment for
each customer.
18
We call this capability, “1-to-1 Customer Engagement”.
1-to-1 Customer Engagement enables companies to transition their marketing away from a
traditional one-to-many campaign-driven approach. A one-to-one approach allows
companies to have consistent, contextual and relevant conversations with individual
customers across any channel or touch point.
The key to achieving 1-to-1 Customer Engagement is an idea that’s simple to conceive, but
very difficult to execute: one centralized brain.
19
In other words, one piece of intelligence that acts as a single decision authority across your
application ecosystem.
Each channel or system profits from this single source of customer intelligence and can
leverage it to gain insights or perform relevant actions.
In Pega Marketing™, this centralized brain is called the Pega Customer Decision Hub, and it
leverages AI to enable 1-to-1 Customer Engagement.
In Pega Infinity™, the Pega Customer Decision Hub forms the core of the customer
engagement platform, which sits at the center of existing systems and channels in an
enterprise.
20
Data from every customer engagement across the enterprise is collected by the Brain and
used to make predictions and decisions about every interaction in every channel.
Continuous learning and decision-making are the foundation of a 1-to-1 Customer
Engagement solution.
The Customer Decision Hub combines analytics, business rules, customer data, and data
collected during each customer interaction to create a set of actionable insights that it uses
to make intelligent decisions. These decisions are known as the Next-Best-Action.
Every Next-Best-Action weighs customer needs against business objectives to optimize
decisions based on priorities set by the business manager.
In the milliseconds before interacting with a customer, the Customer Decision Hub
processes thousands of predictive and adaptive models to determine customer needs,
considering the customer’s immediate context to ensure the Next-Best-Action is relevant,
timely, and contextual. These models can be propensity, risk, or churn models.
Next, the decision strategy considers business rules and matches those with the customer’s
context and higher-level business goals.
21
All of this information is used by the Next-Best-Action decision strategy to evaluate every
potential action that could be taken with a particular customer in a given situation. The
decision strategy then recommends the best way to interact with the customer to achieve
the optimal result.
Using the Next-Best-Action approach, the Customer Decision Hub is able to identify the
best moments for making a sale, providing a service, making a retention offer, or doing
nothing at all (e.g. if nothing is relevant enough to warrant the customer’s attention). Next-
Best-Action is even able to select which offers are most likely to be accepted by the
customer in a sales or retention situation. Next-Best-Action decisions are distributed, in
real-time, to each of your real-time owned channels, such as web, mobile, and contact
center. Through Pega Marketing, Next-Best-Actions can also be distributed to real-time
paid channels such as Google, YouTube, Facebook, LinkedIn and Instagram. Pega
Marketing also integrates with non-real time outbound channels such as data management
platforms (DMPs) and email.
Once the Next-Best-Actions are distributed and customer responses have been received by
the Brain, the whole process begins again, and new Next-Best-Actions are distributed
within milliseconds. Every outbound channel, including a data management platform, is
dynamically updated with the Next-Best-Action to ensure consistency and an optimized
customer experience no matter which channel the customer interacts with.
In summary, the Pega Customer Decision Hub is the Always-On Brain that acts as a single,
centralized decision authority.
22
It uses data about the customer, including past interactions, as input.
It leverages advanced AI techniques to make predictions.
And it uses decision strategies (which combine traditional business rules with predictive,
adaptive and text analytics), to deliver consistent and personalized Next-Best-Actions
across all channels.
23
Next-Best-Action Designer
Introduction
Next-Best-Action Designer guides you through the creation of a Next-Best-Action strategy
for your business. Its intuitive interface, proven best practices and sophisticated underlying
decisioning technology enable you to automatically deliver personalized customer
experiences across inbound, outbound and paid channels. Next-Best-Action Designer is
organized according to the high-level sequence of steps needed to configure the Next-Best-
Action strategy for your organization.
Transcript
Next-Best-Action Designer guides you through the creation of a Next-Best-Action strategy
for your business. Its intuitive interface, proven best practices and sophisticated underlying
decisioning technology enable you to automatically deliver personalized customer
experiences across inbound, outbound and paid channels.
The Next-Best-Action Designer user interface allows you to easily define, manage and
monitor Next-Best-Actions.
The tabs across the top of the user interface represent the steps that need to be completed
to define Next-Best-Actions.
Use the Taxonomy component to define the business structure for your organization.
Use the Constraints component to implement channel limits and constraints.
Use the Engagement policy component to define the rules that control which actions are
offered to which customers.
Use the Arbitration component to configure how actions are prioritized.
Use the Channels component to configure when and where Next-Best-Action is triggered.
24
The system uses these definitions to create an underlying Next-Best-Action Strategy
framework. This framework leverages best practices to generate Next-Best-Action decision
strategies at the enterprise level. These decision strategies are a combination of the
business rules and AI models that form the core of the Pega Centralized Decision Hub,
which determines the personalized set of Next-Best-Actions for each customer.
Use the Taxonomy component to define the hierarchy of Business Issues and Groups to
which an action belongs.
25
A Business Issue is the purpose behind the actions you offer to customers. For example,
actions with the purpose of retaining existing customers should be grouped under the
business Issue of Retention. Actions with the purpose of acquiring new customers belong
to the business Issue of Acquisition.
Business Groups are used to organize customer actions into categories. For example, as
part of the business Issue of Acquisition, you can create Groups for products like Credit
Cards, Mortgages, or Personal Loans, with the intention of offering these to potential
customers.
Use Constraints to specify outbound contact limits as well as to limit overexposure to a
specific action or group of actions.
Customer contact limits allow you to limit the number of interactions that a customer can
receive over a given period of time on a specific channel. For example, you can decide that
you do not want your customers to receive more than two emails per week or two SMS
messages per week.
On the Constraints tab of Next-Best-Action Designer, you can define more extensive
suppression rules by creating Contact Policy rules in the library. Contact Policy rules are
reusable across all Business Issues and Groups.
26
In the Contact Policy library, you define suppression rules that automatically put an action
on hold after a specific number of outcomes are recorded for some or all channels. For
example, an action can be suppressed for a customer for seven days after the customer
has seen an ad for that action five times. Suppressing or pausing an action prevents over-
exposure by limiting the number of times a customer is exposed to the same action.
Use Engagement policies to define when specific actions or groups of actions are
appropriate for customers.
There are four types of engagement policies:
Eligibility determines whether or not a customer qualifies for an action or group of
actions. For example, an action may only be available for customers over a specific age or
living in a specific geographic location.
Applicability determines if an action or group of actions is relevant for a customer at a
particular point in time. For example, a discount on a specific credit card may not be
relevant for a customer who already owns a card.
Suitability determines if an action or group of actions is appropriate for a customer for
ethical or empathetic reasons. For example, a new loan offer may not be appropriate for a
customer whose credit score is low, even though it might be profitable for the bank.
27
Contact Policies determine when an action or group of actions should be suppressed and
for how long. For example, you can suppress an action after a specific number of
promotional messages has been sent to customers. To activate Contact Policy rules created
in the library on the Constraints tab, add them to the Engagement Policy tab.
Arbitration determines how the Customer Decision Hub prioritizes the list of eligible and
appropriate actions that come out of each group.
The factors weighed in arbitration are: Propensity, Context weighting, Action value, and
Business levers, each represented by numerical values. A simple formula is used to arrive
at a prioritization value, which is used to select the top actions.
Propensity is the likelihood of a customer responding positively to an action. This is
calculated by Artificial Intelligence (AI). For example, a click on an offer banner or an accept
of an offer in the contact center are considered positive behaviors.
Real-time contextual data is an important part of making highly relevant
recommendations. Context weighting allows you to assign weighting to a specific context
value for all actions within an Issue or Group. For example, if a customer contacts the bank
to change their address, the weight of the Service context will increase, and the highest
priority action will be to ensure that the relevant service is delivered to the customer.
Action value enables you to assign a financial value to an action and prioritize high-value
actions over low-value ones. For example, promoting an unlimited data plan might be
28
more profitable for the company than a limited data plan. Action values are typically
normalized across Issues and Groups.
Business levers enable you to accommodate ad hoc business priorities by specifying a
weight for an action or Group of actions and/or its associated Business Issue.
Next-Best-Action Designer enables Next-Best-Actions to be delivered via inbound,
outbound and paid channels.
These channels can be toggled on or off. If a channel is toggled off, the Next-Best-Actions
will not be delivered to that channel.
An external real-time channel is any channel that presents actions selected by the
Customer Decision Hub to a customer. These channels can include a website, or a call-
center or mobile application. A real-time container is a placeholder for content in an
external real-time channel.
A trigger is a mechanism whereby an external channel invokes the execution of a Next-
Best-Action decisioning process for specific Issues and Groups. The result will be delivered
back to the invoking channel. For example, when a real-time container called “Mortgages
Landing Page” is configured, the website invokes this real-time container before loading the
mortgage page.
29
As you have seen in this video, Next-Best-Action Designer is organized according to the
high-level sequence of steps needed to configure the Next-Best-Action strategy for your
organization. These steps involve:
• Defining the business structure for your organization
• Implementing the channel limits and constraints
• Defining the rules that control which actions are offered to which customers
• Configuring how actions are prioritized
• Configuring when and where Next-Best-Action is triggered
30
Defining and managing customer actions
31
Action hierarchy
Customer action introduction
In Pega Customer Decision Hub™, next-best-action customer recommendations can take
many forms, such as a banner advertisement, a retention offer, or a service message.
Learn how to recognize different types of actions in your organization and organize similar
actions into action groups.
Actions in your organization
Click the Play icon to learn about customer actions and how to organize them.
Full-screen mode is available in the lower right of the interaction.
In Pega Customer Decision Hub, customer actions are offerings that list the details of your
products. Next best action uses these properties to determine the priority of each offering
for each customer and provide you with the next best action to take.
To learn more about the different type of customer actions, consider the following
scenarios in various channels: contact center, mobile, and web.
32
Click Continue to see how actions are used in various scenarios.
Scenario 1: Contact center: up-sell
Is this scenario a valid customer action?
• True
• False
Correct feedback: This scenario is an upsell offer made in a contact center and is a
valid customer action.
Incorrect feedback: An upsell offer made in a contact center is a valid customer action.
Scenario 2: Contact center: customer service
33
Is this scenario a valid customer action?
• True
• False
Correct feedback: This scenario is a service message recommended by the Pega
Customer Decision Hub and is a valid customer action.
Incorrect feedback: A service message recommended by the Pega Customer Decision
Hub is a valid customer action.
Scenario 3: SMS – retention
34
In your opinion, is this a valid customer action?
• True
• False
Correct feedback: This scenario is a retention offer made on the mobile channel to
retain a customer and is a valid customer action.
Incorrect feedback: A retention offer made on the mobile channel to retain a
customer is a valid customer action.
Scenario 4: Inbound-web: cross-sell
35
Is this a valid customer action?
• True
• False
Correct feedback: This scenario is a banner advertisement on a website and is a valid
customer action.
Incorrect feedback: A banner advertisement on a website is a valid customer action.
Customer action properties
Every customer action has properties that define its characteristics.
In the following image, click the + icons to identify the valid properties of customer actions.
36
Image is a valid property and represents the image that is used for the banner.
Price is a numeric property that represents the price of the customer action.
Benefit is a valid property that represents the value proposition to the customer.
Short title is a valid property that represents a short description of the customer action.
Call to action is not a valid property. This button is a call to action for people who want to
purchase the product.
Company name is not necessarily a valid property, as this text represents the company
name.
Business hierarchy
In Pega Customer Decision Hub, customer actions apply to various business issues and are
organized into a three-level hierarchy. The business issue is the purpose of the actions that
you offer to your customers. Each action that you prepare for your customers is associated
with an issue and a group. Because of the association, you must always define the
37
hierarchy before creating actions.
In the following image, click the + icons to learn more about the three-level hierarchy. Then,
check your knowledge with the following interaction.
Business issue: In Pega Customer Decision Hub, a business issue represents the business
area for which a customer action is applicable (for example, Sales, Retention,
and Service).
38
Group: In Pega Customer Decision Hub, a group organizes actions into logical buckets (for
example, Credit Cards or Mortgages).
Actions: In Pega Customer Decision Hub, an offering is referred to as an action (for
example, Reward card or 30-year fixed rate). Each action that you prepare for your
customers is associated with an issue and a group.
Drag and drop the items from the bottom to correctly complete the business hierarchy.
Correct feedback: That's right! You selected the correct response.
Incorrect feedback: You did not select the correct response. Please try again.
Customer actions and renaming
Pega Customer Decision Hub uses the default term actions to refer to messages delivered
during customer interaction. If needed, you can change the term actions to a more
appropriate term. For example, actions under a Loans group (Sales issue) can be changed
to Loan offers whereas actions under a Loans group (Retention issue) can be
changed to Loan loyalty program.
39
In the center of the following image, slide the vertical line to view a sample hierarchy with
renamed actions. Then, check your knowledge with the following interaction.
Set 1: A customer action is any banner ad, retention offer, or service message.
40
Customer actions are properties that define their characteristics.
Set 2: A customer action can be presented on any customer interaction channel such as
Web, Mobile, Contact center, or Social.
A customer action can be presented only on Web.
Set 3: A customer action is organized into a hierarchical structure. The default hierarchy is:
Business issue -> Group -> Action.
A customer action is organized into a hierarchical structure. The default hierarchy is:
Business issue -> Action -> Group.
Choose the correct statement.
Correct feedback: Well done!
Incorrect feedback: We recommend you go through the material once more.
41
Managing business structure
Introduction
In Pega Customer Decision Hub™, Next-Best-Action Designer is used to build the business
issue/group hierarchies for your organization. Together, business issues and groups form
the organizing structure for your customer interactions. Each Next-Best-Action that is
presented to a customer is associated with a business issue and group.
Transcript
This demo will show you how to set up your business structure in Next-Best-Action
Designer.
This is the Pega Customer Decision Hub™ portal. To manage the business structure,
navigate to Next-Best-Action Designer. Customer actions are organized using a hierarchical
business structure called Issues and Groups. Under the Taxonomy tab, you define the
Issues and Groups that will play a role in the Next-Best-Action decisioning process.
In this case, you will set up U+ bank’s business structure based on their various areas of
business focus. Currently, there is only one Issue in the business structure.
You can add, remove, or create new Issues here by editing and configuring the hierarchy.
Note that, this can be edited only with an additional privilege. Here is a list of Issues that
were created before. You can also create new Issues.
42
Now, add an existing Issue. Similarly, you can add, remove, or create business Groups. To
add more groups, configure the groups.
To complete the configuration, save the changes.
This demo has concluded. What did it show you?
- How to add an Issue to the hierarchy.
- How to add Groups to an Issue.
43
Creating actions
Introduction
An action holds various details about a particular offering, such as Start Date, End Date,
and Expected Revenue. Each action is backed by a decisioning proposition. The system
automatically manages this relationship by creating, deploying, and deleting the
proposition instances as needed. Since an action is closely tied to a proposition, it must
always be created in the context of an issue and group.
Transcript
This demo will show you how to create an Action. It will also explain how to create Actions
in bulk.
U+, a retail bank, recently introduced several new credit card offers that they would like to
display to their customers. As a decisioning consultant, you have been tasked with creating
these offers in the Sales Issue and CreditCards Group as per the pre-defined business
hierarchy.
This is the Pega Customer Decision Hub™ portal. Actions are created and managed on the
Actions landing page.
You can either create a single Action or Actions in bulk.
To create a single Action, enter a short description of the new Action. Select the
appropriate Issue and Group. Now, open the Action.
44
The attributes of an Action distinguish it from other Actions. Attributes are used by the
Customer Decision Hub to select the right Action for a customer. Fill in the attributes
relevant to this Action. For example, provide a Description that explains its purpose and
Benefits that describe how this Action will benefit the customer.
Bear in mind that some of the values you enter may be customer-facing information. For
example, U+ wants to display the content of the Benefits attribute when this Action is
presented on its website. To complete the configuration, save the Action.
To view Actions within a certain Group, you can filter by a specific Issue and Group. This is
the Action you just created. Thus far you have created a single Action. However, sometimes
you may want to create multiple Actions at once. You can do that by uploading a list of
Actions from a Comma Separated Value (CSV) file. To create such a CSV file you need to
download a CSV template and fill it in.
Open the downloaded CSV template. You can use this template to add/update Actions. Add
new Actions with corresponding attributes to the file.
Properties that start with py, such as pyName, pyIssue, pyGroup, pyLabel and
pyIsPropositionActive are mandatory internal properties. It is important that you input
these values in the correct format into the CSV file.
45
The property pyIsPropositionActive corresponds to the Availability property, which is visible
on the Details tab of an Action. Possible values for this property are 'Always', 'Never' and
'Date'.
Choose 'Always' to ensure that the Action is selected during the Next-Best-Action
decisioning process by the Customer Decision Hub.
Set it to 'Never' if you never want the Action to be selected, for example when you want to
retire the Action
If you set the value to 'Date', you should enter a date range in the 'StartDate' and 'EndDate'
fields. The Customer Decision Hub will select this Action only during that time period.
pyIssue corresponds to the Issue the Action was created for.
pyLabel corresponds to a short description of the Action.
pyName corresponds to the Action name.
pyGroup is the Group to which the Action belongs.
Save and close the CSV file.
Now, import Actions from the file. Select the CSV file that contains the list of Actions you
want to import. If you want to delete existing Actions from this Issue/Group that have been
created in the system but not in the CSV file, you can select this option.
For now, you are only interested in adding new Actions.
46
Notice that the window displays a summary of the result of the import.
For example, when a duplicate Action is added, the Errors count is incremented, and you
can download the error file to learn the exact issue. Complete the import. Notice that the
new Actions are now listed on the Actions page.
This demo has concluded. What did it show you?
- How to create an Action.
- How to set Action attributes.
- The most important properties of Actions in a CSV file.
- How to create multiple Actions using a CSV file.
47
Renaming actions
Introduction
By default, Pega Customer Decision Hub uses the term ‘actions’ to refer to messages
delivered during a customer interaction. If needed, you can change this to a more
appropriate term, such as Promotions or Nudges. Actions can be renamed with different
terms based on the business issue and group to which the product belongs.
Transcript
This demo will show you how to change the default naming of Actions to something that
better reflects their business context.
This is the Pega Customer Decision Hub™ portal. On the Actions landing page, you can
view all the Actions that have been created. To view Actions within a certain Group, you can
filter by a specific Issue and Group. In this case, filter Actions in the Sales Issue and
CreditCards Group. Open an Action to view the name. Notice that the Sales/CreditCards
Group uses the default terminology ‘Action’.
U+ bank wants to rename Actions under the Sales Issue and CreditCards Group to better
suit their business purpose. The Actions under the CreditCards Group are always
promotional credit card offers, therefore the bank wants to change the naming convention
from ‘Action’ to ‘Offer’.
You can change the default naming convention of Actions in Next-Best-Action Designer.
To manage the business structure, navigate to the Taxonomy tab of Next-Best-Action
Designer. The ‘Action naming’ convention is set in a business Group’s configuration
settings. This can be set for each group in the business hierarchy.
48
Edit the hierarchy to modify the Group attributes.
In this case, change the Action naming convention to Offer.
To complete the configuration, save the changes. Verify that the Action is renamed. Open
any Action under the Sales Issue and CreditCards Group to view the effect of renaming.
49
From now on, Actions under Sales/CreditCards will be referred to as Offers.
This demo has concluded. What did it show you?
- How to change the default naming convention of Actions, to a name that better
reflects their business context.
50
Presenting a single offer on the web
51
Real-time containers
Introduction
A real-time container is a service that manages communication between Pega Customer
Decision Hub and external channels. An external real-time channel is any channel that
presents actions selected by the Customer Decision Hub to a user or customer. For
example, a website, a call-center application, or a mobile application.
Transcript
This video explains the concept of real-time containers, which manage communication
between the Pega Customer Decision Hub™ and external channels.
This is the website of a retail bank called U+ Bank.
The bank plans to promote new offers on the account page, which is displayed when
customers log in.
52
The marketing department wants to leverage the Customer Decision Hub’s Next-Best-
Action capability to display the right offer for each customer.
The Customer Decision Hub’s real-time container functionality is used to implement this
requirement.
A real-time container is a placeholder for content in an external real-time channel.
An external real-time channel is any channel that presents actions selected by the
Customer Decision Hub to a user or customer. For example, a website, a call-center
application, or a mobile application.
53
Here’s how the website invokes the real-time container to present credit card offers on the
account page.
In the Customer Decision Hub’s Next-Best-Action Designer, a real-time container called
“Account Page Container” is configured. The website invokes this real-time container before
loading the account page.
The Customer Decision Hub then evaluates the actions from the associated Issue/Group,
which in this case is Sales/CreditCards and returns the resulting offer details back to the
website.
54
The website then loads the account page with the content returned by the Customer
Decision Hub, such as the offer image, description and other relevant attributes.
Meanwhile, the Customer Decision Hub records these customer interactions in the
Interaction History. An Impression is recorded to indicate that the action was shown to the
customer, a Click-through is recorded when the customer clicks on the action. Marketers
use these metrics, i.e. Impressions and Click-throughs, to measure the level of customer
engagement, and therefore, the success of the marketing effort.
55
Creating a real-time container
Introduction
Learn how to create and configure a real-time container which allows you to manage
communication between Pega Customer Decision Hub and external channels.
Transcript
This demo will show you how to create and configure a real-time container.
To create real-time containers, navigate to the Real-Time Artifacts landing page.
Enter a short description of the new real-time container.
An external channel uses the real-time container’s identifier when it makes a request to the
Pega Customer Decision Hub.
You can change the identifier to a value that the external channel in your environment
expects.
For example, the U+ Bank website is pre-configured to call a real-time container with the
identifier Ubank_home_banner.
Now, open the real-time container.
Here you can turn the real-time container’s ability to respond to requests from external
channels on or off.
Keeping the default value will allow the Customer Decision Hub to serve requests from the
website.
‘Impression capture’ allows the external channel to control how the Customer Decision Hub
records impressions.
56
‘Captured on retrieval’ means that the Customer Decision Hub records an impression in the
Interaction History immediately after sending the action details to the external channel.
‘Captured by channel’ means that the Customer Decision Hub does not immediately record
an impression. Instead, the external channel can explicitly request that the Customer
Decision Hub records an impression at a later point in time.
In this case, the bank wants the impressions to be captured, so keep the default value.
To record a click in the Interaction History without initiating an action flow for the
customer, keep the default value.
This option is suitable for actions that do not have a defined flow, for example, on the Web
channel when a customer clicking on an ad does not trigger any follow-up steps.
Here you can view the list of Next-Best-Action strategies that you have associated with this
real-time container.
Save the configuration.
Navigate to the Containers tab to see the newly created real-time container.
This demo has concluded. What did it show you?
- How to create a real-time container
- How to configure the real-time container settings
57
58
Presenting a single offer on the web
Introduction
Pega Customer Decision Hub is the always-on customer brain used to select the right offer
to be presented to each customer in any real-time channel. Learn how to configure Next-
Best-Action Designer to select a single offer that will be displayed on a website.
Transcript
This demo will show you how to configure Next-Best-Action Designer to select a single offer
that will be displayed on a website.
U+, a retail bank, would like to use the Pega Customer Decision Hub™ to display a single
offer on its website.
U+ bank wants to make offers related to credit cards and display the same ‘Cash back’ offer
to every customer who logs in to the web site.
For example, if customer Troy logs in to his Accounts page, the ‘Cash back’ offer is
displayed.
If Troy clicks on the ‘Learn more’ button, it takes him to the Offers page.
59
This page shows the offer details.
To implement this business requirement, navigate to Next-Best-Action designer.
Here you can configure business rules to define when specific actions or groups of actions
are appropriate for customers.
Notice that the NBA hierarchy currently has three Business issues with Groups under them.
In this case, U+ wants to promote credit card offers. So, open the CreditCards Group.
You can configure the actions here.
A consultant has already created a few actions under Sales/CreditCards. However, in this
scenario, the bank wants to show only the ‘Cash back’ card.
Now, open on the ‘Cash back’ offer.
To display an offer, you need to add a treatment that is specific to the channel.
In this scenario, U+ bank wants to present the offer on their website, so select the right
treatment type.
60
A web treatment represents the message that you want to communicate visually to your
customer on the web channel. You can either use an existing treatment or create one here.
Now, fill in the required fields for the web treatment.
Here you provide a link to an image that will display the action for this treatment.
Once you provide the link, a preview of the content is displayed.
This is the URL that you want the customer to go to when they click on the action.
61
Now, select the intended location and style of the treatment design. For example, the
treatment can appear as a large central banner, a rotating strip of images, or a footer on
the website.
In this scenario, select Tile to display the ‘Cash back’ offer on the top right of the Account
page.
You can specify the language of the treatment if required. Specifying the treatment
language allows Pega Customer Decision Hub to consider the customer’s preferred
language when selecting the treatment.
Add the web treatment you just created.
In this phase, U+ does not have any further eligibility or prioritization requirements for this
action. Save the changes.
Here, you can enable the channels and triggers that will invoke Next-Best-Actions.
As U+ wants to display the offer on the web, enable the web channel.
Now, configure the real-time container that the U+ website will use to display the offer
banner on the account page.
The real-time container manages communication between the Pega Customer Decision
Hub and external channels such as the web and call center. A decisioning consultant has
already configured a real-time container for you, so select it to be added to your
configuration.
Once the real-time container is added, configure it to select the results from an appropriate
Business issue and Group. In this case, U+ bank wants to display the ‘Cash back’ offer,
which is under Sales/CreditCards.
62
With that, all the necessary configuration for this scenario is complete. Save the changes
for the configuration to take effect.
The account page on the U+ Bank website has been configured to use the real-time
container with the name 'SalesCreditCards' and with Placement type 'Tile'.
The web treatment is shown as a tile on the top right of the page.
When Troy clicks on the ‘Learn more’ button, it takes him to the URL that was configured as
the click-through URL for the treatment.
If another user logs in, they will see the same offer.
This demo has concluded. What did it show you?
- How to configure Next-Best-Action designer to display a web treatment on the U+ bank
website.
63
Defining customer engagement policies
64
Customer engagement policies
Introduction
Engagement policies are a set of business rules and practices used by the organization to
determine which customers are eligible for which Next-Best-Actions. These policies allow
you to specify the conditions under which an action or group of actions are eligible for a
customer.
Transcript
This video explains the concept of customer engagement policies.
The Pega Customer Decision Hub™ combines analytics, business rules, customer data, and
data collected during each customer interaction to create a set of actionable insights that it
uses to make intelligent decisions. These decisions are known as the Next-Best-Action.
Every Next-Best-Action weighs customer needs against business objectives to optimize
decisions based on priorities set by the business manager.
Typically, the business defines a set of rules that make certain actions available to certain
customers. This set of rules is called an engagement policy.
As part of an engagement policy, three types of conditions are defined – Eligibility,
Applicability, and Suitability.
65
Consider the following examples: a retail bank is promoting a Gold Credit Card; a telco is
offering a new iPhone upgrade with an unlimited data plan; and a communications and
media company is promoting a new bundle of HD channels.
Let’s see how to define engagement policy conditions that will ensure the bank’s Next-Best-
Action decisions support these promotions.
In Eligibility, you define strict rules for what is legal, and even possible, to offer
customers. For example, to be eligible for the Gold Card offer, customers must be 18 years
or older.
Similarly, for the iPhone upgrade offer, customers are eligible for a new contract only if
their old contract ends in less than three months.
For the TV channels offer, customers must already own a TV subscription. This offer is not
available for customers who have only mobile or landline subscriptions.
In Applicability, you specify rules for limiting what to offer based on a customer’s current
situation, which is often defined by the products they currently have. These rules are not as
rigid as those for Eligibility.
For example, a Gold Card is not applicable if the customer already has a higher value card,
such as a Platinum Credit Card. If a customer already has a Platinum Card, they might be
eligible for the Gold Card, but the Gold Card is not applicable to them. If they ask for it, they
may get it, but the business would prefer not to present them with the Gold Card offer.
66
Similarly, with the iPhone upgrade offer, if a customer explicitly expressed in the last survey
that they weren’t interested in an iPhone, this action is not applicable to them. For the TV
channels offer, the business does not want to advertise HD channels to a customer who
has recently bought a set top box that is not capable of HD.
In Suitability, you specify conditions that define an offer as appropriate for a customer.
Suitability rules are in place to promote the concept of empathy. That is, to help an
enterprise be empathetic toward their customers and refrain from making offers that may
not be a good fit.
For example, as the Gold Card is a high value card, it is only suitable for a customer whose
debt-to-income ratio is below a certain threshold. Although a customer might be eligible for
it, and the offer might be applicable to them, it would be inappropriate to market it to
them, as there is a risk of default.
Similarly, an unlimited data plan is not suitable to be offered to a customer with low
Internet usage. In the last example, if the customer's favorite TV shows are not available in
HD, then it’s not empathetic to offer them the new HD channels package.
67
U+, a retail bank, has configured its engagement policy to suit its own business objectives
as well as the needs of its customers.
In this scenario, a marketer for U+ has designed 200 actions that could be presented to
customers. To select the Next-Best-Actions from these, the Pega Customer Decision Hub
first checks eligibility conditions and filters the actions. Then, the applicability conditions
are run to filter them further. Next, the suitability conditions are checked to derive the final
set of available actions.
These actions will go through one final stage before being presented to customers: the
arbitration stage. Arbitration is used to prioritize and choose the Next-Best-Actions based
on what is relevant for each customer right now. This is decided by considering factors
such as AI-calculated propensity, the action value, and various business levers.
68
69
Defining eligibility, applicability, and suitability rules
Introduction
Engagement policies are defined in Next-Best-Action Designer. Engagement policies specify
the conditions under which an action or group of actions is available for a customer. These
policies should be defined in the following categories: eligibility, applicability, and
suitability, which represent the true nature of the associated conditions.
Transcript
This demo will show you how to define engagement policy conditions such as Eligibility,
Applicability, and Suitability using Next-Best-Action Designer.
U+, a retail bank, has introduced two new credit cards and would like to offer them to
customers based on certain criteria.
First, all credit cards are eligible only for existing U+ customers who are at least 18 years
old.
In addition, the business wants to ensure that these two new credit cards will be available
for new customers who do not yet have a credit card.
Last, the business understands that not all credit cards are suitable for everyone.
Due to the credit limits of each card, the business wants to offer the Rewards Card to
customers with a credit score higher than 500 and the Rewards Plus Card to customers
with a credit score higher than 750.
This is the Pega Customer Decision Hub™ portal. You define engagement policies in Next-
Best-Action Designer. Engagement policies specify the conditions under which an action or
group of actions is available for a customer.
Now, you will set up the engagement policies to reflect U+ bank’s requirements. The
engagement policies can be defined for a specific group within an issue and/or for
individual actions.
Edit to configure the engagement policies. Notice that the two actions are listed under this
group.
70
As a best practice, engagement policies should be defined in the following categories:
Eligibility, Applicability and Suitability, which represent the true nature of the associated
conditions.
First, define the Eligibility condition to ensure that only current customers are considered
for the action, and that the customers are at least 18 years old.
Then, define the Applicability condition so that only new customers, who currently do not
have a credit card, qualify for the actions. These criteria are also being defined at the group
level.
The "new customer" check is done using the LifeCyclePeriod property. Add a condition
to check if the customer already has a credit card.
71
Lastly, business wants a different Suitability condition for individual cards depending on
the card’s limit. Since the conditions are specific to each credit card, the Suitability
condition for each must be defined at the Action level.
Open the Rewards Card Action. This card comes with a certain credit limit. Therefore, U+
believes this card is only suitable for customers with a credit score higher than 500, even
though customers may satisfy the Eligibility and Applicability conditions. Thus, define the
Suitability condition accordingly.
Next, open the Rewards Plus Card action. This card has an even higher credit limit than the
Rewards card. Thus, this card is suitable for customers with credit score higher than 750.
To complete the configuration, save the changes.
72
Log in as Troy, a 25-year-old Chef, who became a customer of U+ bank 3 months ago, and
his credit score is 600. Notice that the Rewards Card offer is displayed, as Troy satisfies all
Eligibility and Applicability conditions. Troy will only be offered this card because his credit
score is higher than 500 but lower than 750.
Now, login as Barbara, a 40-year-old Engineer, who became a customer of U+ bank a
month ago, and her credit score is 800. Notice that the Rewards Plus Card offer is
displayed, as Barbara also satisfies all Eligibility and Applicability conditions. Barbara on the
other hand will be offered both cards, as her credit score is higher than 750.
73
74
Testing the next-best-action configuration with
persona testing
Introduction
You can design persona-based tests to verify that the Next-Best-Action Strategy gives the
expected results. You can design these personas according to your requirements and the
strategy that you are testing to ensure you do not introduce any regressions as you define
your engagement policies and arbitration.
Transcript
This demo will show you how a persona test is conducted on customer personas to
evaluate the next-best-action strategy results.
U+, a retail bank, wants to verify that customers are receiving the correct offers. In other
words, the bank wants to ensure that the strategy is sending offers to customers in line
with their business requirements.
Personas are a representation of various customer profiles that you use to test the results
of the next-best-action strategy framework.
Persona-based tests use customer personas with specific characteristics to evaluate next-
best-action strategy results.
The bank decides to use Troy, Barbara, and John as personas to test the configurations.
Troy is a 25-year-old chef who became a customer of U+ bank 3 months ago, and his credit
score is 600. Barbara is a 40-year-old engineer who became a customer of U+ bank a
month ago, and her credit score is 800. John is a 45-year-old IT employee who became a
customer of U+ bank 15 days ago, and his credit score is 600.
Persona-based tests are used to verify that the next-best-action strategy gives the expected
results. On the Next-Best-Action Designer Engagement Policy tab, you can create test cases
for any group.
You can create test cases for a specific group, or for all groups. Running test cases for all
groups implies that you are running the next-best-action strategy for all issues and groups
in your business structure.
In this case, as the bank wants to target the credit cards group, let's create a persona test
case at the group level.
75
To configure a persona test case, select the action and treatment that you expect the test
persona to receive according to your next-best-action strategy.
In this case, Troy is supposed to receive the Standard card and Rewards card actions.
In the Persona (Data transform) field, select the persona against which you want to test the
strategy. Select Troy.
In the next-best-action scope section, select whether the test should only check
engagement policy configuration, or include additional elements such as constraints and
arbitration.
Selecting “Engagement policies only” will validate that your policy conditions are providing
the desired results. This will ensure that eligibility, applicability and suitability are tested.
Selecting "Engagement policies and arbitration" will validate the effectiveness of your
policies when arbitrating across all actions. The test will consider arbitration, adaptive
analytics, treatment and channel processing, and constraints.
In this case, select "Engagement policies only".
Once configured, select the test and click the Run selected tests icon.
76
Barbara is a persona who is eligible for the Rewards plus card and the Premier rewards
card. Let's create a test case for this persona.
John is a persona who is eligible for the Standard card and the Rewards card. Let's create a
test case for this persona.
77
Run the test to confirm he passes the test.
The bank has now decided to offer the Standard card to customers who have an average
balance greater than $2000.
So, let's see how this engagement policy change affects these test results. To define an
engagement policy, click Standard card.
Define an Applicability condition so that only customers who have a minimum average
balance of $2000 qualify for the actions.
78
Run all the tests to see how this engagement policy configuration affects the test cases.
Observe that Troy fails the test. This means he does not satisfy all the existing engagement
policies. Analyze the results to see why the test did not give the expected outcome.
Per the report, he is not eligible for the Standard card and, thus, this test failed.
79
Troy’s average balance is 1500, which is less than the defined applicability condition.
Whereas, John’s average balance is 3000 and Barbara's average balance is 3500, which are
more than the defined applicability condition.
The Troy persona failed the test, but John and Barbara passed it.
Let's edit the test for Troy's persona to delete the Standard card assertion and run the test
So, you can design all these personas according to the business requirements and test
them every time an engagement policy has changed.
80
You can also export the results into an excel sheet for further analysis.
This demo has concluded. What did it show you?
- How to create persona test cases.
- How engagement policy changes affect the persona test results.
- How to edit persona test cases.
81
Avoiding overexposure of actions
82
Contact policies
Understanding contact policy requirements
Too many contact attempts over a short period of time can have a negative impact on a
customer's attitude toward further actions by your company. To maximize the lifetime
value of every customer relationship, organizations must prevent outreach fatigue by
optimizing the number of actions taken.
In the Pega Customer Decision Hub, contact policies allow you to suppress actions after a
specific number of outcomes.
Suppressing or pausing an action prevents oversaturation by limiting the number of times
a customer is exposed to the same action.
Defining contact policies
Contact policies determine when and for how long an Action or group of Actions should be
shown to a customer. Contact policies track responses to Actions over a specific period of
time, allowing you to implement rules such as the following:
• Do not show an ad to a customer for two weeks if the customer ignores the ad five
times in a one-week timeframe.
Note: If outcomes are tracked for an individual action, then the action is not shown once the suppression criteria are met.
• Do not show a group of ads for six months if a customer clicks on any ad in the
group 3 times over a period of 30 days.
Note: If outcomes are tracked for all actions in the group, then all of these actions are not shown once the suppression criteria are met.
An Interaction History Summary rule is used to determine the number of impressions and
clicks generated by a customer over a period of time. The default time periods are 7 and 30
days. There might be business requirements to track a customer’s response to an offer
over different time periods, for example, 14 days.
You can add more tracking periods by creating a new Interaction History Summary rule for
the required time period and then updating the part of the Next-Best-Action strategy that
references it.
83
Defining action suppression rules
Introduction
Suppression rules determine when and for how long an action or group of actions should
not be shown to a customer. These suppression rules put an action on hold after a specific
number of outcomes are recorded for some or all channels.
Transcript
This demo will show you how to suppress a single action or group of actions for a limited
time period.
U+, a retail bank, currently displays various credit card offers to each customer who logs in
to the website.
For example, every time Troy logs in to his accounts page, a credit card offer is shown.
Sometimes the same offer is shown multiple times.
For a limited time period, the bank wants to automatically suppress offers that are shown
or clicked too often.
In this scenario, the bank has two requirements. First, do not show a credit card for ten
days if the card was shown three times in the last seven days.
Second, do not show any credit cards for ten days if a user has clicked on a credit card five
times in the last seven days.
Contact policies are used to implement these business requirements. You create contact
policies in Next-Best-Action Designer.
84
On the Constraints tab, you can define the suppression rules by creating contact policy
rules.
For the first requirement, showing a credit card a maximum of three times, configure a
contact policy to track Impressions at the Action level.
Provide a name for the contact policy.
Then select the type of outcome that will be tracked by the contact policy, in this case
Impressions.
You can specify whether the responses are tracked for one specific action, or for all actions
in the group. Track the first requirement at the Action level, since you want to show one
specific card a maximum of three times.
You can select the time period over which the responses should be tracked. In this case,
responses should be tracked over a period of seven days.
The newly configured contact policy has been added.
The first business requirement is to suppress the action for ten days if there are three
impressions for the web treatment, so fill in the details accordingly.
Enter the number of responses required to fulfill the suppression criteria.
85
Select the channel for which the responses are tracked. Note that if you want to track
impressions across multiple channels, you can select Any.
Enter the number of days for which an action should be paused after the suppression
criteria are met.
The next business requirement is to suppress the entire group of actions if there are five
clicks for web treatments.
The first contact policy is configured to track Impressions, so add another contact policy.
For this requirement you will be tracking Clicks for all actions in the group. This because
you want to hide all credit cards if there are five clicks on any one credit card.
Once the contact policy is created, fill in the suppression rule details. If there are five clicks
on web treatments, suppress the action for ten days.
Save the changes.
The contact policy rules are reusable as policy rules across all business issues and groups.
86
As the bank wants to suppress credit card offers, open the CreditCards group.
Add the contact policy rules you just created.
With that, all the necessary configuration for this scenario is complete. Save the changes.
When customer Troy logs in to his accounts page, the Premier Rewards Card offer is
displayed.
After showing him this offer three times, it is automatically suppressed, and a different
credit card offer is shown.
When customer Barbara logs in to her accounts page, a credit card offer is displayed.
After clicking on any credit card offer five times, the credit card offers are not shown again.
This demo has concluded. What did it show you?
- How to define contact policy rules to suppress a single action or group of actions.
87
Arbitrating between actions
88
Action arbitration
Introduction
Pega Customer Decision Hub combines analytics, business rules, customer data, and data
collected during each customer interaction to create a set of actionable insights that it uses
to make intelligent decisions. Arbitration aims to balance customer relevance with business
priorities by weighing numerical values for the following factors: propensity, context
weighting, action value, and business levers. Learn to create a simple formula for arriving
at a prioritization value, which is used to select the top actions.
Transcript
This video explains the concept of action arbitration.
Pega Customer Decision Hub™ combines analytics, business rules, customer data, and data
collected during each customer interaction to create a set of actionable insights that it uses
to make intelligent decisions. These decisions are known as Next-Best-Action.
Every Next-Best-Action weighs customer needs against business objectives to optimize
decisions based on priorities set by the business manager.
U+ Bank, a retail bank, has several actions for its customers and has configured
engagement policies to suit both customer needs and business objectives.
89
In this scenario, a marketer for U+ has designed 200 actions that can be presented to
customers. To select the Next-Best-Actions from these, Pega Customer Decision Hub first
checks the eligibility conditions and filters the actions. Then, the applicability conditions are
run to filter it further. Next, Customer Decision Hub checks the suitability conditions
to derive the final set of available actions.
These actions move through one final stage before being presented to customers: the
arbitration stage. Arbitration is used to prioritize and choose the best actions based on
what is relevant for the customer right now.
Arbitration aims to balance customer relevance with business priorities. The factors
weighed are Propensity, Context Weighting, Action Value, and Business Levers, each
represented by numerical values. A simple formula is used to arrive at a prioritization
value, which is used to select the top actions. The number of top actions selected depends
on the channel of interaction. For example, the top three actions, plus two tiles and
one hero treatment, can be selected for display on a bank’s website.
90
Propensity is the likelihood of a customer responding positively to an action; this is
calculated by AI. For example, the higher the likelihood of a customer accepting an offer,
the higher the Propensity value for that offer.
Context Weighting allows Pega Customer Decision Hub to consider the situational context
for each action. For example, if a customer contacts the bank to close their account, the
highest-priority action is to ensure that the customer is retained. The priority of an action is
increased by a specified value when the context is detected.
Action Value enables you to assign a financial value to an action and prioritize high-value
actions over low-value ones. This value is typically normalized across Issues and Groups.
For example, an unlimited data plan is more profitable than a limited data plan. So, in a
situation where a customer is eligible for both plans, the unlimited plan has higher priority.
Business Levers allow the business to assert some level of control over the prioritization
of actions defined within the system. Levers are used to manually nudge Customer
Decision Hub toward Next-Best-Actions based on external factors. For example, the
recommended Next- Best-Action might be to offer a credit card to a customer when they
visit the home page. But to meet a business goal, the Mortgage Line of Business favors a
mortgage offer even if that offer is ranked a little lower on the list of possible actions.
Consider an example where three actions are selected for arbitration. At the moment, only
the Propensity is used for prioritization.
Action arbitration with propensity before prioritization
The result of the arbitration is that the top action is the one with the highest Propensity.
91
Action arbitration with propensity after prioritization
Examine what happens when Context Weighting together with Propensity are considered
for arbitration. For example, if the intent of a customer calling customer service is to
change their address, the Context Weight of a Service action increases.
Action arbitration with context weight before prioritization
As a result, the Arbitration caters to the current need of the customer and presents a
Service action as the top action for the customer. Thus, the Arbitration caters to the current
need of the customer and presents a Service action as the top action for the customer.
Action arbitration with context weight after prioritization
Consider another scenario in which a customer is eligible for two credit cards and two
other actions. Now, consider that the Action Value is also used in arbitration when
prioritizing. In this case, the Platinum Card is assigned a higher value by the business than
the Gold Card.
92
Action arbitration with action value before prioritization
Thus, the arbitration selects the Platinum Card as the top action.
Action arbitration with action value after prioritization
Finally, consider an example in which all four parameters are used for arbitration. In this
case, U+ Bank wants to promote two new checking account offers under the Sales issue.
The bank sets a higher Business Lever value for the Checking Accounts actions.
93
Action arbitration with business levers before prioritization
Although the Propensity of the Checking Accounts actions is low, they are selected as the
top actions due to their high Lever values.
Action arbitration with business levers after prioritization
94
Action prioritization with AI
Introduction
Explore how AI-based arbitration works and how AI predicts customer behavior. Arbitration
aims to balance customer relevance with business priorities. To select the top actions, a
formula is used to arrive at a prioritization value. The formula uses the propensity value,
which is calculated using AI. Propensity is the predicted likelihood of positive behavior,
such as the likelihood of a customer accepting an offer.
Transcript
This demo will explore how AI-based arbitration works and explain how AI predicts
customer behavior.
U+, a retail bank, uses the Pega Customer Decision Hub™, to display marketing offers to
customers on its website. The bank would like to display more relevant offers to customers
based on their behavior.
Troy, a customer, qualifies for two credit card offers. When he logs into the bank’s website,
he sees the top offer for him, the Standard Card.
These are the Arbitration settings defined in Pega Customer Decision Hub’s Next-Best-
Action Designer. Arbitration aims to balance customer relevance with business priorities.
To achieve this balance, Propensity (P), Context weighting (C), Action value (V), and Business
levers (L) are represented by numerical values and plugged into a simple formula, P * C * V
* L. This formula is used to arrive at a prioritization value, which is used to select the top
actions.
95
Notice that only Propensity is currently enabled. Propensity is the predicted likelihood of
positive behavior, such as the likelihood of a customer accepting an offer. The value of
Propensity is calculated using AI.
Note the Propensity and Priority values of the Standard Card. The Propensity for every
action starts at 0.5 or 50%, the same as the flip of a coin. This is because in the beginning,
the AI has no past customer behavior on which to base its predictions. Propensity is one of
the factors used to arbitrate between relevant offers and select the top offer for a
customer.
Notice that although only Propensity is enabled for arbitration, the value of Priority, which
is currently based on Propensity only, does not match the Propensity value. This is because
the Priority calculation doesn’t use the raw Propensity value directly. Instead it uses the
value resulting from a built-in Propensity smoothing mechanism. The Propensity
smoothing mechanism is used to jump-start the process of AI learning. It helps to equalize
the sudden changes in Propensity values calculated by AI during the initial phase of its
learning, when it has yet to gather enough customer behavior data to make accurate
predictions.
96
If Troy doesn’t click on the current offer this time, a different offer will be shown the next
time he visits the website. The next offer Troy is eligible for, the Rewards Card, is then
selected for display. If Troy ignores this card as well, by not clicking on it, then the next time
he logs in, the Standard Card offer will be displayed again. Why this behavior? First, Troy
only qualifies for these two credit card offers. Second, the AI model behind these offers is
configured to treat an Impression as a negative behavior. In other words, when a customer
is presented with an offer but doesn’t click on it, the AI records this as a negative behavior.
As a result, the Propensity, and therefore the Priority, of the not-clicked-on offer decreases.
Notice that the Propensity value of the Standard Card offer dropped from 0.5 to 0.25.
97
Now, if Troy clicks on the ‘Learn more’ link for the Standard Card offer, a positive response
is recorded, and thus the Propensity value of the Standard Card increases.
98
The Customer Decision Hub is configured to calculate the Propensity for each Treatment.
To understand how this works, let’s examine the AI behind a Treatment. This pop-up
window provides a summary of the AI behind this Treatment. In the Pega Customer
Decision Hub, the AI that determines the Propensity for positive behavior towards an
action or Treatment is called an adaptive model. From here, you can navigate to the
adaptive model itself.
99
An adaptive model is a self-learning predictive model that uses machine learning to
calculate Propensity scores. It automatically determines the factors that help in predicting
customer behavior. These predictors can include a customer’s demographic details,
product and service usage, past interactions with the bank, and even contextual
information such as the current channel of interaction.
This adaptive model considers an Impression, when a marketing offer is displayed on a
website, a negative behavior. It considers a Click a positive behavior.
100
Therefore, when a customer sees an offer message but doesn’t click on it, the model
records a negative behavior.
The monitoring tab provides an overview of the model’s performance. The business can
use this information to assess the contribution of the model’s predictions with respect to
the success of the actions.
The model report provides more insight into the AI model itself. This AI model is
automatically generated by the system, and it adapts its prediction algorithm in real-time,
based on incoming customer responses. The report shows more information about the
predictors, such as how they are grouped and details a data scientist can use to analyze the
current health of the model and diagnose any potential problems.
In the Predictor report, you can examine the performance of individual predictors. Let’s
examine the LifeCyclePeriod predictor.
This a predictor of type Symbolic. The individual Predictor report shows that a customer
whose lifecycle stage is RETAIN is most likely to accept the Standard Card action in the web
channel.
101
The behavior of one customer can influence the Propensity calculation for other customers
with a similar profile. For example, when Robert, a customer with a profile similar to Troy,
logs in, he is shown the same offer as Troy. The same AI model is behind the Treatments
for both customers, so Robert’s action will influence Troy’s Propensity score.
102
Prioritizing actions with business levers
Introduction
Often, due to an internal ad-hoc priority, the business would like to boost the chance of
certain actions being selected. To achieve this, they would like to present more relevant
offers to customers based not only on their behavior but also on business priorities. Learn
how to include business requirements in an action prioritization calculation to boost the
chance of an action being selected.
Transcript
This video will show you how to include business requirements in an action prioritization
calculation to boost the chances of an action being selected.
U+, a retail bank, noticed that one of the offers, the Rewards card offer, was not presented
frequently enough due to its low propensity because customers ignored it during the initial
launch.
For example, Troy, a customer, qualifies for two credit card offers – the Standard Card and
the Rewards Card. When he logs in to the bank’s website, he sees the top offer for him,
Standard Card.
Now, due to an internal ad-hoc priority, the bank wants to boost the chances of the
Rewards Card being selected as the top offer. That is, the bank would like to present more
relevant offers to customers based on not only their behavior but also on business
priorities.
103
To implement this requirement, you must first enable the Action Weighting, a Business
Lever, in Next-Best-Action Designer’s Arbitration equation. This ensures that an action’s
business weight is used in the priority value calculation.
In this case, the bank wants to boost the Rewards Card. So, open the Rewards Card offer.
Edit the offer to set a business weight, a value in percentage, that is required to boost the
offer. In this case, U+ wants to increase the changes of this action selected by 10%.
Save the offer for the changes to take effect.
Now, when Troy logs into the website, he will see that Rewards card is the top offer.